Download Repurchase agreements and the law

Repurchase agreements and the law: how legislative changes fueled the housing bubble Fiona Maclachlan Manhattan College Riverdale, NY 10471 [email protected] Abstract J.R. Commons maintained that the fundamental unit of analysis in institutional economics is the transaction, which is characterized by three aspects: conflict, dependence and order. The order is established through rules, whether customary or legal. In contrast to a general equilibrium framework, institutional economics considers the rules establishing the order to be endogenous, evolving alongside other socio-­‐economic variables. Commons’ idea of a transaction is particularly useful in the analysis of inventive financial contracts for which the rules that establish the order are subject to interpretation by the courts, and in some cases, to legislative change. As the rules change, so do their effects on the wider economy. In this paper, the recent evolution of the law with respect to repurchase agreements is examined. Repurchase agreements (repo) are short-­‐term debt contracts that were central in the expansion of liquidity during the run-­‐up to the financial crisis. The irresponsible and disruptive lending within the housing market was related to excessive financial institution leverage made possible through repo. A key issue in the law relating to repo is the status of the collateral in the event of bankruptcy of the borrower. The development of this aspect of the law is traced through interpretations of the bankruptcy code by the courts, as well as through legislated changes to the code. In the latter, evidence of vested interests tilting the law away from the public interest is found. Keywords: Endogenous Credit; Institutional Economics; J. R. Commons; Repurchase Agreements; Repo; Securities Financing; Shadow Banking; Automatic Stay; Safe Harbor Provisions JEL Classification Codes: B52; E51; G01; G21; G2 ______________________________________________________________________________________ For presentation at the annual meeting of the Association for Evolutionary Economics, Philadelphia, PA; January 3-­‐5, 2014. Preliminary draft: comments are welcome. Commons (1931) argues that the distinguishing feature of institutional economics is its emphasis on the transaction, rather than preferences and production functions, as the fundamental unit of analysis. According to Commons, a transaction involves three aspects: conflict, dependence and order; where order refers to the rules, either legal or customary, governing the transaction. The recent financial crisis points to the relevance of Commons’ approach. This paper illustrates the relevance with an analysis of the evolving set of rules governing the short-­‐term collateralized credit instruments known as repurchase agreements, or repo. I argue that a change legislated as part of the Bankruptcy Abuse Prevention and Consumer Protection Act (BAPCPA) of 2005 contributed to the fragility of the financial system in the years running up to the crisis. Furthermore, I note that the change in the law is evidence of the detrimental effect of the powerful financial lobby, as the change benefited financial market participants at the expense of the efficiency and stability of the wider economy. Bankruptcy Law and the Automatic Stay Provision Bankruptcy protection for failing businesses is an important institution that serves to bring order to what otherwise would be a chaotic process of creditors competing for their share of the assets of an insolvent firm. For a business, a commonly used alternative to a Chapter 7 liquidation is a Chapter 11 reorganization. Under Chapter 11, the bankruptcy court arranges a reorganization of failing firm’s business including a restructuring of the firm’s debts. The aim is to allow the firm to continue to operate, thereby generating maximum benefits for all stakeholders. The continued operation of the firm is beneficial to shareholders, employees, customers, suppliers, and even communities in the case of large firms. Since the aim of the reorganization is to preserve as much economic value as possible, the creditors as a group will normally be able to recover more than they would under Chapter 7; although, fully secured creditors often have more to gain from an immediate liquidation (Warren 2008, 62). 2 A key feature of bankruptcy protection is the imposition of automatic stay. The rule of automatic stay freezes creditors’ claims when a firm files a bankruptcy petition, and, where a voidable preference can be shown, forces creditors to return assets collected during the period immediately preceding the bankruptcy filing (Warren 2008, 98). The obvious benefit of the automatic stay is that it facilitates the troubled firm’s efforts to generate revenue and maximize its economic value. The firm is not forced to liquidate assets to meet the demands of creditors, and can use available cash flows for operating expenses. A less obvious benefit of the automatic stay is that, because it shifts some of the costs of the bankruptcy to the creditors, the creditors have greater incentives to prevent one from occurring. First, they have an incentive not to lend to bankruptcy-­‐prone firms; and secondly, in the event that they have lent to such a firm, they have an incentive not to take actions that could push the firm over the edge into bankruptcy protection. These two incentive effects of automatic stay are important in transactions involving financial firms in which the collateral is fairly liquid, and adjustable through margin calls. Without the automatic stay, the focus of the lender shifts from the credit-­‐worthiness of the borrower to the market value of the collateral. Furthermore, without automatic stay provisions, a decline in securities prices would more likely lead creditors to inundate troubled firms with demands for more collateral, forcing them to sell assets at distressed prices and increasing the likelihood of a bankruptcy filing. With the automatic stay, it is in the interest of the creditor to keep the debtor out of bankruptcy to avoid having their claims tied up in the bankruptcy proceedings. Application of Automatic Stay to Repo A sale and repurchase agreement (or repo, for short) is a contract in which one party sells an asset and, at the same time, makes a promise to buy it back after a specified period. The period can be as short as one day. The price at which the asset is exchanged is usually less than its market value. This discount is expressed in percentage terms and is referred to as the haircut, or margin. A repo is similar to a secured, or collateralized, loan because if the original seller fails to repurchase the 3 securities, the counterparty is entitled to retain ownership. But an important difference between the securities in a repo transaction and collateral in a regular secured lending transaction is that the repo contract may give the lender additional rights over the collateral. For example, the lender may be allowed to sell the securities that were delivered and return comparable ones. Another feature that can be included in the repo contract is permission for the lender to re-­‐use, or rehypothecate, the securities in order to borrow in the repo market from another counterparty. These features of the contract suggest that repo is not exactly equivalent to a secured loan, but is rather a sui generis hybrid between a securities sale and a secured loan (Stigum 1989, 332). The hybrid nature of repo has, in the past, led to uncertainty as to whether the automatic stay should be applied if a repo borrower files for bankruptcy. In the late 1970s and early 1980s, there were a number of failures of smaller securities dealers due to losses from repo. Two prominent examples, both from 1982, were the failures of Drysdale Government Securities and Lombard-­‐Wall. In the case of Lombard-­‐Wall, the court ruled that the securities that had been purchased in the repo transactions were not the property of the repo lenders and thus, were subject to automatic stay. The ruling was contrary to the expectations of the market participants who thought repo should be construed a sale rather than a loan. Stigum (1989, 219) quotes a market participant who said the ruling “scared the hell out of the industry.” As a result of the Lombard-­‐Wall ruling, an effort was mounted to get an amendment to the bankruptcy code that would exempt repo collateral from automatic stay. The Treasury department, however, did not endorse the change to the code, expressing the opinion that the exemption from automatic stay would reduce the incentive of repo lenders to lend only to sound institutions (Roe 2011, 567). Despite the position expressed by the Treasury department, the bankruptcy code was amended in 1984 to exempt from automatic stay in repo the following types of collateral: Treasury and federal agency securities, bank certificates of deposit, and bankers’ acceptances. 4 These instruments have come to be known as traditional collateral, while non-­‐
traditional collateral refers to other types of securities used in repo that did not have granted in 1984, the explicit exemption from automatic stay. As the market for private label mortgage backed securities (MBS) and asset backed securities (ABS) grew, so did the repo market in which they served as collateral. According to Acharya and Öncü (Acharya et al. 2010, 330), repo financing was extended to this riskier, non-­‐traditional collateral beginning in the mid-­‐90’s. Repo desks at the broker-­‐dealers found that that making repo loans with non-­‐traditional collateral was profitable, and seemed to be low risk. It helped the securitization part of the firm’s business to be able to offer repo financing to buyers. It was similar to an automobile company having a financing arm to help consumers purchase cars. Since repo was short-­‐term, the exposure to losses was limited. If a counter-­‐party looked shaky, the repo desk could increase the margin or simply refuse to roll over the repo. In the worst case scenario of a default, the lender could seize the collateral and sell it. If a sufficient margin had been applied, the chance of losses was minimal. Demand for repo financing also came from the proprietary trading operations of the broker-­‐dealers, as well as from hedge funds and mortgage real estate investment trusts (mREITs). The supply, or funding, side came money market funds including unregistered liquidity funds, as well as securities lenders who discovered they could use the repo market to invest the cash collateral left with them. The Fed’s low interest rate policy in the early 2000s created an incentive for money market funds to venture into riskier repo, as a means of generating enough revenue to cover their operating costs. Still another factor responsible for the growing use of non-­‐
traditional collateral was the shortage of traditional collateral that was emerging as a result of the demand for it in derivatives transactions and in payments systems (Gorton 2010, 23). Since non-­‐traditional collateral did not have the exemption from automatic stay, repo contracts were written to represent the transaction as a true sale, so that if the 5 borrower filed for bankruptcy, the lender could retain possession of the securities. In 2000, however, a court ruling in the bankruptcy of Criimi Mae, a publicly held commercial mortage REIT, took market participants by surprise, creating a disturbance in the repo market (Schroeder 2002, 568). Against the expectations of the market, the court ruled that the repo Criimi Mae was using to finance its assets was equivalent to a secured loan and that the automatic stay should be applied to the collateral. The case was similar to Lombard-­‐Wall. In both instances, the lenders found a bankruptcy court had thwarted their efforts to represent repo as something other than collateralized lending. And in both cases, an effort was mounted to get an amendment to the bankruptcy code to exempt repo collateral from automatic stay. The chief argument for an exemption for repo collateral, as well as collateral used in derivatives transactions, was that a stay could create financial gridlock. A huge volume of cash flows through complex daisy chains of transactions every day. If the flow is stopped at any point, the whole system could seize up. Managing cash inflows and outflows is challenging in the best circumstances. The position was that if collateral was tied up in bankruptcy proceedings, a systemic crisis could ensue. The arguments of the financial lobby were successful and the change in the code to exempt non-­‐traditional collateral from the automatic stay was made as part of the Bankruptcy Abuse Prevention and Consumer Protection Act (BAPCPA) of 2005. In contrast to the period twenty years earlier when changes to the bankruptcy code relating to repo collateral were debated, the argument made by market participants for an expansion of the safe harbor from automatic stay went largely unnoticed. For example in the dissenting and minority views represented in the Report of the Committee of the Judiciary House of Representatives that accompanied BAPCPA, there was no mention of the automatic stay exemptions. Role of Automatic Stay Exemption in the Housing Bubble While financial gridlock is a serious concern, it is not clear that changes to the bankruptcy code are the best way to deal with it, when other strategies, such as 6 expanding the lender of last resort function of the central bank, are available. The problem with the exemptions from automatic stay, as the Treasury argued in the early 1980s, is that they weaken the incentives of lenders to forgo lending to risky, bankruptcy-­‐prone borrowers. The expected result is more bankruptcies, leading to grater economic instability and wasteful liquidations. The housing bubble that preceded the financial crisis of 2007-­‐2008 has many possible causes and it will be some time before economists reach a consensus on which were most important. But the role of the change in the bankruptcy law to exempt repo collateral from automatic stay deserves further attention. It has been mostly neglected in the economics literature, although there have been a number of law journal articles on the issue, for example, Lubben (2010) and Roe (2011). To see how an apparently minor change in the bankruptcy code could fuel a housing bubble in which even homebuyers who are likely to default on their mortgages are able to get access to funds, one could follow a typical financing chain and consider how it would be different in the absence of an exemption from automatic stay for repo collateral. A subprime mortgage lender will typically get the funds to pass along to the homebuyer by selling the mortgages to an investment bank. The investment bank then pools the mortgages and creates its own private label MBS that it can sell to hedge funds and mREITs, or keep on its own balance sheet. Since risky mortgages carry higher interest rates, the securities carry high yields. Endemic to this system of financing is the widely recognized principal-­‐agent problem in which the managers of the investment banks, hedge funds, and mREITs have an incentive to value short-­‐term over the long-­‐term profits. But the problem is amplified with the possibility of using repo to create leverage to scale up the size of the investments in the private-­‐label MBS. Without repo the size of the market for private label MBS would be constrained, because in order to purchase more of these risky assets, the investors would either have to raise equity, or access longer term credit, which is more expensive and less plentiful. The repo lenders are willing to part with current dollars to be fed into the financing chain, because they assume the 7 asset they are getting in return, a fully secured overnight loan, to be liquid and safe. The increased use of repo financing for mortgage assets was evident at Lehman Brothers subsequent to the change in the bankruptcy code. Lehman grew its mortgage related assets 66% from 2006 to 2007, and by the first quarter of 2008 had $197 billion in repo loans outstanding (Angelides and Thomas 2011, 326). And the strategy was not unique to Lehman Brothers. Brunnermeier (2009, 80) notes that the “fraction of total investment bank assets financed by overnight repos roughly doubled from 2000 to 2007.” If, however, the repo collateral did not have the exemption from automatic stay the repo lenders would be more reluctant to participate in the market, especially with bankruptcy prone counterparties, like leveraged hedge funds, and investment banks with aggressive proprietary trading operations. Without the dollars flowing in from the repo lenders, the amount of leverage and spending on high yield MBS would have been constrained. And without the spending on the private label MBS, there would have been less demand for the risky mortgages that went into them. Furthermore, without the exemption from automatic stay, lenders would have had less incentive to hasten the decline of shaky borrowers by refusing to roll over the short-­‐term repo, or by demanding additional margin. Conclusion By taking a Commons approach of paying attention to the rules that govern transactions, we have uncovered an aspect of the current financial system that has arguably increased its fragility. The bankruptcy code evolved to contain in it the automatic stay provision, which, along with the rule of voidable preference, works to create a situation in which the interests of the debtor and creditor are aligned towards keeping the debtor afloat. It works first by discouraging creditors from extending loans to bankruptcy-­‐prone debtors-­‐-­‐no matter how valuable the collateral; and secondly, by discouraging creditors from taking actions like demanding more collateral or refusing to roll over short term loans that could push 8 a troubled debtor into the arms of the bankruptcy court. Automatic stay has the effect of creating more dependence in the creditor-­‐debtor relationship, and results in a more stable financial system. But the automatic stay is not in the interest of the financial institutions that participate in the repo and derivatives markets, and it is natural that they would lobby to have the bankruptcy code amended to grant exemptions to those contracts. The creditors want the freedom and flexibility to exit quickly when their counterparties get into trouble. And the debtors understand that removing the automatic stay gives them access to inexpensive short-­‐term credit that otherwise would be more difficult to obtain. The response to the financial crisis has been stricter regulation to contain the excesses of the credit markets. The problem with this approach is the ability of financiers to invent instruments and structures that fall outside of the scope of the regulations. The beauty of the automatic stay provision for repo collateral is that the financial industry was unable to get around it by inventing new contracts, which is the reason that had to get the law changed. Changing the law back would be a straightforward way to reduce the financial instability engendered by repo financing. 9 References Acharya, Viral V., Thomas F. Cooley, Matthew P. Richardson, and Ingo Walter. 2010. Regulating Wall Street: The Dodd-­‐Frank Act and the New Architecture of Global Finance. John Wiley & Sons. Angelides, Phil, and Bill Thomas. 2011. The Financial Crisis Inquiry Report: Final Report of the National Commission on the Causes of the Financial and Economic Crisis in the United States (Revised Corrected Copy). Government Printing Office. Brunnermeier, Markus K. 2009. “Deciphering the Liquidity and Credit Crunch 2007-­‐
2008.” The Journal of Economic Perspectives 23 (1): 77–100. Commons, John R. 1931. “Institutional Economics.” American Economic Review: 648–657. Gorton, Gary. 2010. Slapped by the Invisible Hand: The Panic of 2007. Oxford: Oxford Univ. Press. Lubben, Stephen J. 2010. “Bankruptcy Code without Safe Harbors,” American Bankruptcy Law Journal 84: 123. Roe, Mark J. 2011. “The Derivatives Market’s Payment Priorities as Financial Crisis Accelerator.” Stanford Law Review 63 (3) (March 1): 539. Schroeder, Jeanne L. 2002. “Repo Opera: How Criimi Mae Got Repos Backwards, A.” Am. Bankr. LJ 76: 565. Stigum, Marcia L. 1989. The Repo and Reverse Markets. Homewood, Ill.: Dow Jones-­‐
Irwin. Warren, Elizabeth. 2008. Chapter 11: Reorganizing American Businesses : Essentials. New York, NY: Aspen Publishers. 10